Radar-Aided Vehicle Odometry for Position and Heading Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle odometry systems face challenges in maintaining accurate position and heading estimation due to sensor data variability and environmental uncertainties, which affect the precision of vehicle control in advanced driver-assistance systems (ADAS).
Innovation Solution
A radar-based method is employed to adapt and enhance odometry filters using radar data, correcting initial vehicle positions and headings through an adaptive Kalman filter, improving accuracy by integrating radar data with IMU, WSS, and SAS data to calculate corrected longitudinal, lateral, and heading estimates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional sensor data (IMU, WSS, SAS) is used for odometry estimation, then the system structure is simple, but measurement precision deteriorates due to sensor data variability and environmental uncertainties
Solution Approach 1:
The patent combines radar data with traditional sensor data (IMU, WSS, SAS) in an adaptive Kalman filter framework. The radar provides orthogonal velocity and relative position measurements that are fused with inertial and wheel speed data to correct odometry estimates, thereby improving measurement precision through data merging.
Solution Approach 2:
The adaptive Kalman filter uses radar measurements to compute position and heading errors, which are then fed back to correct the odometry estimates. This feedback mechanism continuously adapts the filter parameters based on radar data quality, improving estimation accuracy while managing system complexity through intelligent parameter adaptation.
2Measurement precision
If radar data is integrated to correct odometry estimates, then measurement precision improves, but device complexity increases due to additional sensor integration and adaptive filtering
Solution Approach 1:
The Kalman filter parameters are made adaptive rather than fixed. The system dynamically adjusts filter gains based on radar data quality and environmental conditions, allowing the measurement precision to improve while the complexity increases are managed through parameter adaptation rather than structural complexity.
Solution Approach 2:
The patent changes the parameters of the Kalman filter based on radar measurements and error calculations. By adapting filter parameters (such as process noise covariance and measurement noise covariance) based on actual performance, the system achieves improved precision without requiring fundamentally more complex system architecture.
3Reliability
If multiple sensors (IMU, WSS, SAS, radar) are integrated, then reliability of vehicle control improves, but device complexity increases
Solution Approach 1:
The adaptive Kalman filter serves as a universal processing framework that handles multiple sensor types (IMU, WSS, SAS, radar) through a unified mathematical model. This multi-functional approach improves reliability by systematically fusing data from all sensors while managing complexity through a single versatile algorithm rather than separate processing chains.
Data Source
AI summary
A method for odometry estimation includes receiving a first sensor data and a second sensor data. The first sensor data is generated by an inertial measurement unit (IMU) of the vehicle, a wheel speed sensor (WSS), and a steering wheel angle sensor (SAS) of the vehicle. The second sensor data is generated by a radar of the vehicle. The method further includes determining an initial longitudinal position, an initial lateral position, and an initial heading of the vehicle using the first sensor data. The method further includes determining a longitudinal position error, a lateral position error, and a heading error of the vehicle using the second sensor data. Moreover, the method includes correcting the initial longitudinal position, the initial lateral position, and the initial heading of the vehicle using the longitudinal position error, the lateral position error, and the heading error.